1 Calibrating Function Points Using Neuro-Fuzzy Technique Vivian Xia NFA Estimation Inc. London, Ontario, Canada Danny Ho IT Department.

Slides:



Advertisements
Similar presentations
FPA – IFPUG CPM 4.1 Rules.
Advertisements


NCSR “DEMOKRITOS” Institute of Nuclear Technology and Radiation Protection NATIONAL TECHNICAL UNIVERSITY OF ATHENS School of Chemical Engineering Fuzzy.
Metrics. A Good Manager Measures measurement What do we use as a basis? size? size? function? function? project metrics process metrics process product.
ESCOM/April 2001Aristotle University/Singular Int’l1 BRACE: BootstRap based Analogy Cost Estimation Automated support for an enhanced effort prediction.
Neuro-Fuzzy Algorithmic (NFA) Models and Tools for Estimation Danny Ho, Luiz F. Capretz*, Xishi Huang, Jing Ren NFA Estimation Inc., London, Ontario, Canada.
Software project management (intro)
1 PROJECT SIZING AND ESTIMATING - EFFECTIVELY USING FUNCTIONAL MEASUREMENT Southern California Software Process Improvement.
A Neuro-Fuzzy Model with SEER-SEM for Software Effort Estimation Wei Lin Du, Danny Ho*, Luiz F. Capretz Software Engineering, University of Western Ontario,
Software Metrics/Quality Metrics
1 Prediction of Software Reliability Using Neural Network and Fuzzy Logic Professor David Rine Seminar Notes.
1 U08784 Software Project Management lecturer: Timothy Au url:
Copyright © The David Consulting Group, Inc. 1 UNDERSTANDING and EFFECTIVELY USING FUNCTIONAL MEASUREMENT Presented By The David Consulting Group.
Software Engineering: A Practitioner’s Approach
Cmpe 589 Spring Software Quality Metrics Product  product attributes –Size, complexity, design features, performance, quality level Process  Used.
FAULT DIAGNOSIS OF THE DAMADICS BENCHMARK ACTUATOR USING NEURO-FUZZY SYSTEMS WITH LOCAL RECURRENT STRUCTURE FAULT DIAGNOSIS OF THE DAMADICS BENCHMARK ACTUATOR.
Software Project Estimation IFPUG VS COSMIC
Artificial Neural Networks (ANN). Output Y is 1 if at least two of the three inputs are equal to 1.
Chapter 6 : Software Metrics
Function Point Analysis What is Function Point Analysis (FPA)? It is designed to estimate and measure the time, and thereby the cost, of developing new.
Software Cost Estimation 1. APPROACHES Traditional: LOC estimation Modern: Functional Point Analysis 2.
1 UseCase-based effort estimation of software projects TDT 4290 Customer-driven project IDI, NTNU, 14. Sept Bente Anda, Simula Research Lab., Oslo,
Software Engineering SM ? 1. Outline of this presentation What is SM The Need for SM Type of SM Size Oriented Metric Function Oriented Metric 218/10/2015.
UKSMA 2005 Lessons Learnt from introducing IT Measurement Peter Thomas –
1 Estimation Function Point Analysis December 5, 2006.
Lecture 4 Software Metrics
Pfleeger and Atlee, Software Engineering: Theory and PracticePage 3.1 © 2006 Pearson/Prentice Hall 3.1 Tracking Progress : Gantt Chart Activities shown.
Estimating Software Projects & Activity Scheduling in the Dynamic, Multi-Project Setting: Choosing Heuristics Through Deterministic Simulation.
Function Point Analysis. Function Points Analysis (FPA) What is Function Point Analysis (FPA)? Function points are a standard unit of measure that represent.
SEG3300 A&B W2004R.L. Probert1 COCOMO Models Ognian Kabranov.
Introduction to Software Project Estimation I (Condensed) Barry Schrag Software Engineering Consultant MCSD, MCAD, MCDBA Bellevue.
Akram Bitar and Larry Manevitz Department of Computer Science
Estimating “Size” of Software There are many ways to estimate the volume or size of software. ( understanding requirements is key to this activity ) –We.
1 Software Engineering: A Practitioner’s Approach, 6/e Chapter 15a: Product Metrics for Software Software Engineering: A Practitioner’s Approach, 6/e Chapter.
Random Forests Ujjwol Subedi. Introduction What is Random Tree? ◦ Is a tree constructed randomly from a set of possible trees having K random features.
Advanced Software Engineering Lecture 4: Process & Project Metrics.
NASA Software Assurance Symposium 2001 Metrics for Fault-Tolerant Real-Time Software Afzel Noore Computer Science and Electrical Engineering West Virginia.
FUNCTION POINT ANALYSIS & ESTIMATION
Software Cost Estimation By Deepika Chaudhary. In project planning process a project is split into a number of activities which are performed in parallel.
The article written by Boyarshinova Vera Scientific adviser: Eltyshev Denis THE USE OF NEURO-FUZZY MODELS FOR INTEGRATED ASSESSMENT OF THE CONDITIONS OF.
1 2  Why SNAP? (in a nutshell) ?  What is SNAP (in a nutshell) ?  Does SNAP work?  How to use SNAP when we already use Function Points?
Data Mining: Concepts and Techniques1 Prediction Prediction vs. classification Classification predicts categorical class label Prediction predicts continuous-valued.
A Presentation on Adaptive Neuro-Fuzzy Inference System using Particle Swarm Optimization and it’s Application By Sumanta Kundu (En.R.No.
VIDYA PRATISHTHAN’S COLLEGE OF ENGINEERING, BARAMATI.
Software Engineering, COMP201 Slide 1 Software Engineering CSE470.
Cost9b 1 Living with Function Points Bernstein and Lubashevsky Text pp
Estimation Questions How do you estimate? What are you going to estimate? Where do you start?
Cost23 1 Question of the Day u Which of the following things measure the “size” of the project in terms of the functionality that has to be provided in.
Prediction of Interconnect Net-Degree Distribution Based on Rent’s Rule Tao Wan and Malgorzata Chrzanowska- Jeske Department of Electrical and Computer.
Internal Logical Files (ILF) An internal logical file (ILF) is a user identifiable group of logically related data or control information maintained within.
THE FAMU-CIS ALUMNI SYSTEM
Testing Techniques.
RET Rules One of the following rules applies when counting RETs:
FUZZY NEURAL NETWORKS TECHNIQUES AND THEIR APPLICATIONS
Sizing With Function Points
Function Point Analysis
Software Development & Project Management
Mk II Function Point Analysis
Prof. Carolina Ruiz Department of Computer Science
Software Size Measures for Cost Estimation
Function Point.
Chapter 5: Software effort estimation- part 2
Software Metrics “How do we measure the software?”
network of simple neuron-like computing elements
More on Estimation In general, effort estimation is based on several parameters and the model ( E= a + b*S**c ): Personnel Environment Quality Size or.
COCOMO Models.
Fuzzy Logic Colter McClure.
Software Effort Estimation
Prof. Carolina Ruiz Department of Computer Science
COCOMO MODEL.
Presentation transcript:

1 Calibrating Function Points Using Neuro-Fuzzy Technique Vivian Xia NFA Estimation Inc. London, Ontario, Canada Danny Ho IT Department HSBC Bank Vancouver, BC Canada Department of Electrical and Computer Engineering University of Western Ontario London, Ontario, Canada Luiz F Capretz

2 Roadmap Concepts of Calibration Neuro-Fuzzy Function Points Calibration Model Validation Result Conclusions

3 Calibration Concept DET, RET --- Component Associated Files Same methodology for all FP 5 components Data Element Types (DET) Record Element Types (RET) Low Average 2-5LowAverageHigh 6+AverageHigh Internal Logical File (ILF) Complexity Matrix External Input, External Output, External Inquiry Internal Logical File, External Interface File

4 Calibration Concept Cont’d e.g. One project has 3 Internal Logical Files (ILF) ILF AILF BILF C DET RET333 Original ClassificationAverage Low Original Weight Value10 7 Observation 1 Ambiguous Classification Observation 2 Crisp Boundary Calibrate complexity degree by fully utilizing the number of component associated files Calibrate to fit specific application

5 Calibration Concept Cont’d ComponentLowAverageHigh External Input 34 6 External Output 45 7 Internal Logical File External Interface File External Inquiry Calibrate UFP weight values to reflect global software industry trend Unadjusted Function Points Weight Values UFP weight values are determined in 1979 based on Albrecht’s study of 22 IBM Data Processing projects

6 Neuro-Fuzzy Function Points Calibration Model Overview Project Data Calibrated by Fuzzy Logic Calibrated by Neural Network Validated for better estimation Estimation Equation ISBSG 8 MMRE, PRED

7 Calibrating by Fuzzy Logic Fuzzy SetFuzzy Rule Fuzzy Inference Output Input Fuzzy Logic System

8 Calibrating by Neural Network Learn UFP weight values by effort the values should reflect complexity complexity proportioned to effort 15 UFP inputs as neurons Back-propagation algorithm

9 Data Source --- ISBSG Release 8 ISBSG International Software Benchmarking Standards Group Non-profit organization Release 8 Contains 2,027 projects 75% built in recent 5 years Filter on ISBSG 8 data set Filter Criteria: Quality, Counting method, Resource level, Development Types, UFP breakdowns Shrink to 184 projects

10 Validation Methodology Developed a calibration tool Randomly split data set totally 184 data points 100 training points 84 testing points for validation Repeat 5 times Using estimation equation for comparison

11 Validation Results (MMRE) Exp.1Exp.2Exp.3Exp.4Exp.5 MMRE Original MMRE Neuro-Fuzzy IMPRV % 20% 19%25%26%22% Avg. IMPRV % 22% MMRE: Mean Magnitude of Relative Error Criteria to assess estimation error The lower the better

12 Validation Results (PRED) PRED: Prediction at level p Criteria to assess estimation ability The higher the better

13 Conclusions Neuro-Fuzzy Function Points model improves software cost estimation by an average of 22%. Fuzzy logic calibration part improves UFP complexity classification. Neural network calibration part overcomes problems with UFP weight values.